Two-dimensional gel electrophoresis (2D-gel) is a technique capable of separating thousands of polypeptides on a single gel. Its usefulness is limited, however, by the difficulty in biochemically identifying the proteins separated on the gels, especially the minor ones. We are engaged in developing a methodology that, if successful, will make it possible to use 2D-gel to select for cDNA clones that code for specific polypeptides. The basic approach is to first isolate mRNA from rat liver epithelial (RLE) cells, reverse transcribe it and clone the subsequent cDNA into a lambda-ecc bacteriophage library designed so that recombinants can be positively selected. The cDNA is inserted between a T7 promoter and terminator which allows it to be transcribed in vitro to produce ersatz mRNA which is, ideally, an accurate representation of the genuine mRNA from which the library is derived. The complexity of the library can be accessed by running 2D-gel on the translation products of the ersatz mRNA. Clones for particular polypeptides can be isolated by sequentially subdividing the library and reducing its complexity until only a single clone responsible for a specific spot is left. We have compared the 2D-gel patterns from genuine mRNA isolated from non-trans- formed RLE cells and cells transformed by the v-Ha-ras oncogene. The patterns are very similar, but some consistent differences were noted, particularly two spots of approximately 15 kDa in the ras transformed cells. Lambda-ecc cDNA libraries have been constructed and work is currently focused on conditions for the accurate expression and transla- tion of the ersatz mRNA.